When Averages Fail: Commentary on The Self-Insurer’s August Issue

In the August edition of The Self-Insurer, Bruce Shutan published “Reinforcing the Crystal Ball of Care,” pulling together perspectives on how predictive analytics is reshaping self-funded healthcare. It’s a thoughtful piece, and worth your time, but reading it as a stop-loss underwriter or actuary raises a sharper question: are we prepared for how fundamentally predictive analytics will disrupt the way we price and manage risk?

Because here’s the uncomfortable truth: actuarial manuals and traditional rating tools may check the regulatory boxes, but they don’t equip you to compete, or win, in today’s volatile stop-loss market.

The Problem with Averages

Bruce’s article captured the promise of predictive tools, but from the trenches of stop-loss, the issue comes down to this: averages don’t underwrite reality.

Two members with identical demographics and ICD codes will never behave the same way. One quietly stays below spec. The other spirals into a million-dollar claim. Yet legacy models keep flattening these trajectories into the same predicted cost. 

The Shift That’s Already Happening

The industry players Bruce highlighted illustrate how predictive analytics is evolving:

• Claros Analytics gives consultants volatility models to stress-test “what if” scenarios.

• Leap Health applies predictive insights to site-of-care, redirecting costly specialty infusions to lower-cost, more patient-friendly settings.

• Prealize Health advances predictive modeling through academic research, pushing accuracy at the theoretical level.

Together, these approaches demonstrate the breadth of innovation underway. But what underwriters need most isn’t just theoretical accuracy or point solutions. They need actionable insights that plug directly into underwriting.

That’s where Merit Predict comes in. By acquiring ground-up claims data on demand and applying advanced AI at the member- and group-level, we provide underwriters with precise risk signals in time to actually use them. This isn’t incremental. It’s underwriting reimagined for the speed and complexity of today’s market.

Why Stop-Loss Can’t Wait

This isn’t academic. For underwriters, the ground is already shifting:

• More inputs such as pharmacy feeds, wearables, and digital health data are flooding in.

• On-demand scenario testing is becoming table stakes.

• Predictive integration with care management will turn underwriting from a pricing exercise into a lever for outcomes.

• Stability is no longer a pipe dream; it’s achievable when groups are priced with actual precision.

The actuary who leans on static tools risks being left behind. The one who embraces predictive precision is shaping the future.

The Takeaway

Bruce Shutan’s article rightly underscores the momentum in predictive analytics. But from the perspective of stop-loss, the real story is starker: clinging to averages erodes profitability; embracing precision builds sustainable books and strengthens the employer market.

Merit Medicine is proud to be part of this transformation; bringing predictive insights out of the academic journals and into the daily workflows of stop-loss actuaries and underwriters.

The actuarial manuals will get us know half way there. Predictive precision will get us the rest of the way.

You can read the full article by Bruce Shutan in The Self-Insurer (August 2025 issue, pp. 22–31).